Infection is a major cause of morbidity and mortality in cancer patients. Our studies are devoted to developing methods to define patients at high risk for infection, improving the ability to diagnose infections early, treat them effectively and ultimately prevent them. We are seeking to differentiate granulocytopenic patients at heightened risk for infection by measuring tissue-bound phagocytes, opsonizing antibody and local defense factors. We are seeking to improve rapid diagnosis of fungal infections by measuring circulating antigens and to define the role of viruses in infection in immunocompromised patients. Effective management of the granulocytopenic patient requires empiric broad-spectrum antibiotic therapy. We are developing new therapeutic approaches based on new antibiotic developments, particularly the third generation cephalosporins. Our results show that a new cephalosporin, ceftazidime, is as effective as a triple drug combination. Our studies are also defining the appropriate antibiotic therapy for patients with unexplained fever, the choice of empiric antifungal therapy, and the merits of empiric therapy vs invasive procedures in patients with pulmonary infiltrates. Prevention of infection has largely resided in decreasing acquisition of new organisms and in reducing the endogenous burden with prophylactic antibiotics. We are continuing our study of total protected isolation for high risk patients but are developing new prevention strategies to improve host defenses. These include passive immunization with post-vaccine antisera to the core glycolipid of enterobacteriaceae as well as pooled immunoglobulins. Our preclinical studies are further focused on attaching (arming) leukocytes with polyclonal or monoclonal antibodies to improve their bactericidal activities and ultimately leukocyte transfusion therapy. We are also developing methods to accelerate granulopiesis using chemically-defined immunoregulatory agents, and ultimately to study their mechanisms of action.